Substance use disorders (SUDs) are highly comorbid with posttraumatic stress disorder (PTSD; Kessler et al., 1995), and this particular comorbidity is associated with a greater likelihood of substance use treatment failure (e.g., Ouimette et al., 1998). Further, research suggests that it is not traumatic exposure per se that puts substance users at greater risk for treatment failure, but whether or not PTSD is present (Chilcoat & Breslau, 1998). Despite some evidence that there is something unique to the presence of PTSD that negatively affects the course and outcome of substance use treatment, studies have not yet comprehensively examined whether this heightened risk is due to current PTSD, past PTSD, or simply a consequence of severe past traumatic exposure. Further, studies have yet to explore the specific mechanisms underlying this increased risk. Mechanisms may be identified from the self-medication model of substance use (Brady et al., 2004). Building from this model, SUD patients with current PTSD may be at risk for treatment drop-out to the extent to which they exhibit emotional vulnerabilities in the form of (a) emotion deregulation; (b) anxiety sensitivity; (c) distress intolerance; (d) emotional avoidance; and/or e) HPA axis dysfunction. To test the mediational role of these variables in treatment drop-out, we will examine four groups (n = 50/group) of inner-city drug users in a residential treatment center: (1) SUD patients with no history of DSM-IV PTSD criterion A traumatic exposure; (2) SUD patients with a history of DSM-IV PTSD criterion A traumatic exposure but not meeting criteria for current or past PTSD; (3) SUD patients with a history of DSM-IV PTSD criterion A traumatic exposure not meeting criteria for current PTSD (past month), but meeting criteria for past PTSD (criteria met for a PTSD diagnosis before one month ago); and (4) SUD patients meeting criteria for current PTSD (which includes the presence of criterion A traumatic exposure). We expect that SUD patients with current PTSD, as compared to all other groups, will exhibit the highest rates of treatment drop-out. Further, emotion dysregulation, anxiety sensitivity, emotional avoidance, low distress tolerance, and HPA axis dysfunction will mediate this relationship, even when controlling for variables often associated with treatment failure (e.g., demographics, negative affect, treatment readiness, severity of past traumatic exposure, level of drug dependence, psychiatric comorbidity). [unreadable] [unreadable] [unreadable]